The AIDS pandemic has increased the need to gather comprehensive data on sexual behavior in developing countries. If, however, reporting of sexual activity is unreliable, our ability to monitor and evaluate AIDS prevention programs is undermined. Moreover, inaccurate reporting of sexual behavior affects explanations about the underlying mechanism driving the epidemic and provides a misleading picture of HIV/STI risk. In addition to not knowing who is having sex, under what circumstances and with whom, estimates of condom use are apt to be seriously biased. Not only is accurate self reporting critical for household surveys that document risky sexual behavior, it is also essential for clinical trials that investigate the effectiveness of technologies to prevent transmission of HIV and other STIs, e.g. female controlled microbicides, the female condom and the diaphragm. Given the explicit expectations that participants comply with project protocols during a clinical trial, accurate data on these behaviors may be difficult to obtain. Furthermore, since researchers asking about new technologies are likely to be perceived as favoring them, acceptability data on new reproductive technologies are thought to be compromised because of courtesy bias. Clearly, improving self-reports of sexual behavior and product use will minimize the possibility that a drug or technology would be considered ineffective when in fact low compliance, perhaps caused by low acceptability, is the issue. Through a series of methodological experiments and innovative research activities, this project aims to (1) assess and improve the validity and reliability of reporting of sexual behavior in survey research implemented in developing countries, (2) improve the quality of behavioral data collected in clinical trials of products and technologies aimed at reducing STI transmission and (3) evaluate the use of new technologies for collecting data on sexual behavior in developing countries, specifically computerized self-interviewing.